Our group is extremely well poised to make a substantial contribution to GTEx, having developed many of experimental and statistical methods for epigenetics now in general use. We discovered CpG island shores, regions of greater disease and population variation than the CpG islands themselves, as well as large hypomethylated blocks. We also pioneered efforts to identify both tissue-specific differentially methylated regions, or t-DMRs, and population variable methylated regions, or VMRs. Finally, we have been pioneers in the integration of methylation, SNP, and expression data to define disease-relevant phenotypes. Of particular importance to this application, we have led efforts to apply these methods on a population level. In this application, we take a strategic approach primarily focused on mental health relevant brain regions, but also on the general question of tissue variation. Our focus on mental health is because of all human diseases these remain particularly opaque to understanding due to the inherent subjectivity of clinical examination; as well as the value of our recently developed methods for deconvoluting the neuron- and glia-specific signals to total DNA methylation and to differential methylation, of great importance for the brain but also to GTEx generally. In Aim 1, we will use whole genome bisulfite sequencing (WGBS) to identify brain region-specific VMRs as well as t-DMRs in four regions of particular importance to three devastating mental health disorders, schizophrenia, depression, and addiction: namely anterior cingulate cortex, frontal cortex, hippocampus, and nucleus accumbens. We will study a sufficient number of samples (30 each from matched patients) to identify both t-DMRs and VMRs. Identification of VMRs requires this larger sample number, since by definition they are variable within a brain region, across individuals. We have shown that t-DMRs are relatively poor indicators of VMRs, but it is the VMRs that are more likely to be related to genetic variation in the population. We will complement this analysis with hydroxymethylcytosine sequencing of the same brain regions because of the potential importance of this modification in the brain. We will also include 5 additional GTEx tissues that exist in reasonable homogeneity compared to many others in the resource. These will serve as controls, for example to compare variability between brain regions to variability across tissues, and also for the purpose of additional t-DMR identification. In Aim 2, we will perform capture bisulfite sequencing to identify brain region-specific VMRs in the population, studying a larger number of individuals (100), in order to relate differential methylation to genotype and gene expression, and targeting the VMRs and t-DMRs identified in Aim 1. We will also perform preliminary analysis of the relationship of VMRs and genetic variation and gene expression, using the existing GTEx data on expression and genetic variation. The work proposed here will provide a critical resource for investigators studying psychiatric disease, and will provide a robust platform and datasets for relating gene expression, methylation, and DNA sequence across populations.